Alerting's Position in the Three Pillars of Observability

Alerting's Position in the Three Pillars of Observability An architecture diagram generated by Archify. Metrics · CPU, memory, requests · Observability Metrics CPU, memory, requests Logs · event records · Observability Logs event records Traces · distributed request flow · Observability Traces distributed request flow Backend rules · or derived metrics · Alerting · focal Backend rules or derived metrics focal Notifications · Slack, Email, SMS, PagerDuty · Alerting Notifications Slack, Email, SMS, PagerDuty Escalation · configured on-call service · Alerting Escalation configured on-call service evaluate notify escalate Observability Alerting Legend Alert rules engine Observability pillar Notification channels Responders

Three pillars

  • • Metrics: quantitative state of the system (CPU, memory, request count)
  • • Logs: detailed records of events
  • • Traces: request flow in distributed systems

Responsibilities

  • • Metrics, logs and traces are common signals, not an exhaustive list
  • • A backend must support the rule and its input
  • • Delivery and escalation need configured integrations